Does Meta retargeting still work in Morocco after iOS ATT?

Yes, provided you've migrated to server-side CAPI. On client Pixel alone, you lose 15 to 20% of iOS conversions and 30 to 45% of iOS members from custom audiences. With CAPI + enriched EMQ (hashed email and +212 phone), Meta matching rebounds from 52% to 88-94%, which rebuilds 25 to 35% of the lost signal. Engagement audiences on Page, IG and video remain completely immune to ATT — that's the fallback when the Pixel suffers.

Retargeting audience window: 30 days or 90 days in Morocco?

Depends on the buying cycle. B2B and premium services (law firms, private schools, real estate): 30 to 90 days depending on commercial cycle — typically 60 days median in Morocco. E-commerce under 500 MAD AOV: 7 to 30 days, past which the audience cools and ROAS collapses. E-commerce above 1,500 MAD AOV: 30 to 60 days. Webotic rule of thumb: keep a hot 7-14 day audience (product page, Checkout) and a warm 30-day audience for mass retargeting — past 60 days, split into a reactivation audience with different creative.

Does Meta dynamic catalog retargeting work for a Moroccan e-commerce account?

Yes, and it's the single most profitable retargeting weapon for Moroccan e-commerce in 2026. Prerequisites: product catalog synced via Shopify or WooCommerce feed, content_id perfectly aligned across site, Pixel, CAPI and feed, ViewContent and AddToCart events fired via CAPI with content_ids. On Webotic accounts, DPA ROAS lands 2.4 to 3.1x higher than static-creative retargeting on the same audience. The post-ATT fix is to feed the catalog server-side rather than relying on a leaky client Pixel stream on iOS.

How many retargeting audiences should a Moroccan SMB run in parallel?

For 8,000 to 30,000 monthly visitors: 4 to 7 audiences total stacked into two retargeting ad sets at most (one hot 14 d, one warm 30 d), with a hierarchical exclusion stack to prevent overlap. Below 8,000 visitors / month, a single fused ad set, otherwise audiences drop below Meta's learning-phase critical mass. Above 30,000 visitors, add a third ad set "DPA catalog 4-14 d" for e-commerce. Above 80,000 visitors, segment finely by cart value or category.

What's the ideal Meta retargeting frequency cap in Morocco?

3 to 4 impressions per user over a rolling 7 days, per audience. Past that, CTR drops below 0.8%, CPM creeps up, and negative comments start landing on the creatives. Set the cap via Reach objective when feasible, or via an exclusion rule: drop anyone with 4 impressions in the last 7 days. On CBO or Advantage+ Audience accounts, cap management is exclusion-based only — Meta doesn't expose a native campaign-level cap in those modes.

What's the minimum budget to run retargeting seriously in Morocco?

Retargeting needs a budget proportional to audience size. For a Moroccan SMB with 8,000 to 15,000 monthly visitors, plan 2,500 to 4,500 MAD/month dedicated to retargeting — 25 to 35% of total Meta budget (itself floored at 8,000 MAD to stay in the learning phase). Below 1,500 MAD/month on retargeting, frequency becomes unpilotable — you either over-impression a small audience or under-impression and never exit learning.

How do you cleanly exclude existing customers from retargeting audiences?

Sync your CRM (HubSpot, Pipedrive, Shopify) with Meta via auto-update Custom Audience, hashed on email and +212 phone. Webotic's data.webotic.ma platform runs that sync continuously. Then exclude that CRM audience from every retargeting ad set, on a 60-day minimum window for e-commerce (median Moroccan repurchase cycle > 60 d) and 90 to 180 days for B2B accounts depending on commercial cycle. Commonly forgotten exclusion that removes 12 to 22% of budget waste on scaling accounts.

Back to blog

/ 10 · retargeting-meta-casablanca-rabatMeta CAPIMay 12, 202613 min

META · POST-ATT RETARGETING

The Meta retargeting that still works in Casablanca and Rabat after iOS ATT.

iOS ATT gutted half of every Moroccan advertiser's Pixel audience. What's left: server-side audiences built through CAPI, Page-IG-video engagement, dynamic catalog. Figures pulled from 28 Meta accounts run by Webotic between January and May 2026.

By Marouane MohachtouFounder & CEO

15-20%iOS ATT lossPixel conversions

25-35%CAPI recoveryserver-side signal

−62%Retargeting CPLvs cold · Webotic median

3-4 / 7 dFrequency capper retargeting audience

What iOS ATT actually broke for Meta retargeting in Morocco

Since App Tracking Transparency rolled out on iOS in April 2021, the mechanics of Meta retargeting have changed for good. In Morocco, iOS share sits at 25-30% of web traffic overall but climbs above 40% on premium segments — luxury, new-build real estate in Anfa, private schools in Rabat. Across the 28 accounts Webotic runs, the Pixel alone now loses 15 to 20% of reported conversions compared with a clean CAPI server-side setup. The bigger damage isn't in the dashboard, though — it's in the audiences themselves. A vanilla "site visitors, 30 days" audience built purely on Pixel ends up missing 30 to 45% of its iOS members who never opted in to ATT. That audience becomes mechanically smaller, more expensive to scale, and Meta starts squeezing frequency to 8-10 impressions per user per week — creative-fatigue territory with bloated CPMs. The second-order effect is nastier: custom audiences built on granular events (Add to Cart, Initiate Checkout, ViewContent on product pages) lose precision because the iOS events only partially fire client-side. Meta ends up learning on a biased subset of your traffic, which also drags down any lookalike layered on top. The pragmatic 2026 verdict: never build a retargeting audience on Pixel alone again. CAPI server-side is the baseline, not a bonus.

Morocco iOS share Q1 2026: 25-30% of web traffic, 40%+ on premium segments.
ATT opt-in in Morocco: 18% average, below the MENA mean of 22%.
Pixel-only 30-day visitor audience: 30 to 45% of iOS members missing.
Pathological frequency observed on Pixel-only: 8-10 impressions / 7 days.

Server-side audiences via CAPI: the technical floor of 2026 retargeting

Server-side tracking through Meta CAPI — ideally orchestrated by GTM Server-Side — fundamentally changes Moroccan retargeting. Your events (PageView, ViewContent, Lead, Purchase) fire from your server-side GTM with enriched EMQ parameters: hashed email, hashed +212 phone, first/last name, first-party customer ID. Meta can then match those events to real Facebook and Instagram accounts even when the iOS client blocks browser tracking. On Webotic accounts that have migrated, Meta's matching rate jumps from 52% (client Pixel only) to 88-94% (Pixel + CAPI + enriched EMQ). That rebuilds 25 to 35% of lost signal — not the full amount, but enough for custom audiences to become operational again. Three server-side audiences to always carve out for Casablanca and Rabat. First, "30-day site visitors, CAPI events only" — the broad, warm baseline. Second, "product page / pricing page visitors, 14 days" — strong-intent, highest-converting. Third, "Initiate Checkout without Purchase, 7 days" for e-commerce, or "Lead form open without submit, 14 days" for B2B / lead gen. Operational detail: don't deduplicate Pixel and CAPI by killing the Pixel — keep both running in parallel with a shared event_id. Meta handles dedup itself, and the dual signal is what maximises matching quality.

Pixel client only: 52% match · Pixel + CAPI + enriched EMQ: 88-94%.
Hash +212 phone in full format before send (sha256, no spaces, no + or 00).
Keep Pixel + CAPI parallel with shared event_id for Meta-side dedup.
Baseline audience: 30-day visitors via CAPI · hot audience: 7-day product page.

Page, IG and video engagement audiences: the retargeting iOS can't touch

Engagement audiences built inside Meta itself — interactions with your Facebook page, your Instagram account, your videos, your Lead form opens, your IG Shop — don't depend on the Pixel or on ATT. Meta compiles them from in-platform data, so iOS has zero leverage on them. In Morocco, where per-user social engagement runs structurally 2 to 2.5x higher than European averages (longer Instagram and Facebook time-on-app, higher per-post interaction on retail and hospitality verticals), these audiences are big and qualitative. Five engagement audiences to keep live in parallel on a serious Casablanca or Rabat account. (1) "Facebook Page engagers, 365 days" — profile visits, post interactions, messages. (2) "Instagram business profile engagers, 365 days" — including Reels views and story interactions. (3) "VV75 video views, 90 days, all videos" — proven-interest audience. (4) "Lead form opened without submit, 90 days" — only for accounts running Lead Ads, declared intent. (5) "IG Shop engagers, 180 days" — for e-commerce accounts that have switched on IG Shopping. Across the 28 Webotic accounts, these five stacked audiences average 3.2x the size of the "30-day site visitors" audience. They keep retargeting alive when the Pixel suffers and give breathing room during seasonal spikes where site-custom audience cost climbs (Black Friday, last two weeks of Ramadan).

Meta engagement audiences: immune to iOS ATT and ad-blockers.
Stacked size observed: 3.2x the 30-day site visitor audience.
Morocco engagement: 2 to 2.5x higher per user than European average.
Standard windows: 365 d for Page/IG, 90 d for VV75, 90 d Lead form.

Dynamic catalog retargeting: what still works (and what doesn't)

Dynamic Product Ads (DPA) remain the single most profitable e-commerce retargeting weapon in Morocco, but the build has shifted post-ATT. What still works in 2026: a product catalog synced via Shopify or WooCommerce feed, wired to Meta with a shared event_id between Pixel and CAPI, feeds both Advantage+ Shopping and DPA retargeting properly. On Webotic accounts running DPA cleanly, catalog ROAS lands 2.4 to 3.1x higher than static-creative retargeting on the same audience. What no longer works the way it used to: "ViewContent per product ID" audiences built on Pixel alone, which depend on a granular event stream that iOS partly drops. The fix: feed the catalog server-side (ViewContent and AddToCart events via CAPI with content_ids) instead of leaning on the client Pixel stream. Three operational rules for DPA in Morocco. First, map the catalog with content_id perfectly aligned across site, Pixel, CAPI and product feed — any drift and Meta refuses to serve dynamic creatives. Second, segment catalog retargeting by window: 1-3 days for hot abandoned carts, 4-14 days for product views, 15-30 days for the cold funnel with different creative messaging. Third, exclude buyers from the last 60 days across all DPA audiences — Moroccan e-commerce AOV is around 380 MAD and the median repurchase cycle exceeds 60 days, so shorter exclusions burn conversion, longer ones burn budget on already-served customers.

DPA retargeting ROAS: 2.4 to 3.1x higher than static-creative retargeting.
Feed the catalog via CAPI (server-side content_ids), not the Pixel alone.
Segment by window: 1-3 d cart · 4-14 d product view · 15-30 d cold funnel.
Exclude 60-day buyers (median MA e-commerce repurchase cycle > 60 d).

Creative rotation, frequency caps, exclusions: the operational discipline

A high-performing retargeting audience in Morocco burns fast if the operational discipline slips. Three rules to enforce from week one. First, creative rotation. On retargeting, never run fewer than 3 or more than 5 active creatives per ad set. Below 3, fatigue kicks in within 5 days (CPM 1.5-2x, CTR halved). Above 5, learning disperses and you can't read per-variant performance anymore. Webotic cadence: 2 to 3 new creatives injected weekly into retargeting, with rest cycles for any asset whose 7-day rolling frequency tops 3.5. Second, frequency caps. The recommended cap for Morocco: 3 to 4 impressions per user per 7 days, per audience. Past that, CTR drops below 0.8% and CPC inflates. Set the cap at the campaign level via the Reach objective when feasible, or via an automatic ad-set-budget cut rule when frequency exceeds 4. For CBO + Advantage+ Audience accounts, cap management is exclusion-based: pull anyone with 4 impressions in the last 7 days out of the retargeting audience. Third, active exclusions. Every retargeting audience must exclude: (1) buyers / converted leads on the right window — 30 d minimum B2B, 60 d minimum e-commerce; (2) employees and internal team via custom audience email — the exclusion everyone forgets, polluting stats and frequency; (3) Careers and Contact page visitors with no commercial intent — noise that inflates audience size without buying intent. A maintained exclusion matrix cuts retargeting CPL by 18 to 28% over a 60-day window.

3 to 5 active creatives per retargeting ad set, never fewer, never more.
Frequency cap 3-4 / 7 d per audience · exclude past 4 impressions.
Mandatory exclusions: 60-day buyers, employees, Careers/Contact visitors.
Maintained exclusion discipline: −18 to −28% retargeting CPL over 60 d.

Retargeting vs cold CPL/CPA in Morocco: what the real numbers say

Across 28 Webotic-piloted Moroccan Meta accounts in Q1 2026 with a complete CAPI server-side setup and a maintained exclusion matrix, retargeting delivers a median CPL 62% lower than cold prospecting. By vertical: on a B2B account (Casablanca law firm, private Rabat school, accounting firm), the cold median CPL sits at 75 MAD, retargeting CPL at 28 MAD — −63%. On an e-commerce account with a 380 MAD AOV, cold Purchase CPA is 110 MAD, retargeting Purchase CPA is 38 MAD — −66%. On a new-build real estate account in the Casablanca-Rabat axis, cold form CPL is 65 MAD, retargeting form CPL is 22 MAD — −66%. The gap only narrows on two verticals: local hospitality (retargeting CPL −38%, because the site-visitor audience is small and social engagement does most of the work) and B2C webinar training (retargeting CPL −47%, because the cold lookalike is already strong). Median Webotic retargeting ROAS in 2026 reaches 3.4x on scaling accounts and 4.8x on mature accounts with full server-side attribution, versus 1.6 to 2.0x on cold. That asymmetry justifies allocating 30 to 45% of total Meta budget to retargeting — a share most Moroccan advertisers don't reach (Webotic measures 12-18% on average across accounts audited pre-engagement).

B2B retargeting CPL: 28 MAD vs 75 MAD cold (−63%).
E-commerce retargeting Purchase CPA: 38 MAD vs 110 MAD cold (−66%).
New-build real estate retargeting CPL: 22 MAD vs 65 MAD cold (−66%).
Recommended retargeting budget share: 30-45% · current MA benchmark: 12-18%.

How many parallel audiences, and how to avoid overlap

The same question lands in every Webotic audit: how many retargeting audiences should we run in parallel in Morocco without saturating segments or burning budget on overlap? The pragmatic answer for a Moroccan SMB with 8,000 to 30,000 monthly visitors: 4 to 7 active audiences total, organised into two retargeting ad sets at most, with a hierarchical exclusion stack. A template that holds up well in Casablanca-Rabat. Hot retargeting ad set: "Product / pricing page visitors 14 d CAPI" ∪ "Initiate Checkout / Lead form open 14 d" ∪ "IG profile engagers 30 d". Excluded: 60-day buyers, Careers page visitors, employees. Warm retargeting ad set: "30-day site visitors CAPI" ∪ "Facebook Page engagers 90 d" ∪ "Video VV75 90 d" ∪ "IG Shop engagers 90 d". Excluded: everyone in the hot ad set + 60-day buyers. That hierarchy structurally eliminates overlap — a hot visitor never gets served by the warm ad set, no double impression over 7 days, frequency stays under control. Past 30,000 visitors / month, peel off a third ad set "DPA catalog 4-14 d" for e-commerce. Past 80,000 visitors / month, segment by cart value or product category. Below 8,000 visitors / month, a single fused retargeting ad set is enough — otherwise audiences fall below Meta's learning-phase critical mass. The standing rule: fewer audiences, better hierarchised, with strict exclusions, always beats ten audiences overlapping and burning budget bidding against themselves on the same users.

Typical Moroccan SMB: 4 to 7 audiences total, 2 retargeting ad sets.
Hot ad set (14 d product/pricing/Checkout) excluded from warm ad set (30 d site).
Below 8,000 visitors/month: one fused retargeting ad set, otherwise too small.
Above 80,000 visitors/month: segment by cart value or product category.

FREQUENTLY ASKED

Does Meta retargeting still work in Morocco after iOS ATT?: Yes, provided you've migrated to server-side CAPI. On client Pixel alone, you lose 15 to 20% of iOS conversions and 30 to 45% of iOS members from custom audiences. With CAPI + enriched EMQ (hashed email and +212 phone), Meta matching rebounds from 52% to 88-94%, which rebuilds 25 to 35% of the lost signal. Engagement audiences on Page, IG and video remain completely immune to ATT — that's the fallback when the Pixel suffers.
Retargeting audience window: 30 days or 90 days in Morocco?: Depends on the buying cycle. B2B and premium services (law firms, private schools, real estate): 30 to 90 days depending on commercial cycle — typically 60 days median in Morocco. E-commerce under 500 MAD AOV: 7 to 30 days, past which the audience cools and ROAS collapses. E-commerce above 1,500 MAD AOV: 30 to 60 days. Webotic rule of thumb: keep a hot 7-14 day audience (product page, Checkout) and a warm 30-day audience for mass retargeting — past 60 days, split into a reactivation audience with different creative.
Does Meta dynamic catalog retargeting work for a Moroccan e-commerce account?: Yes, and it's the single most profitable retargeting weapon for Moroccan e-commerce in 2026. Prerequisites: product catalog synced via Shopify or WooCommerce feed, content_id perfectly aligned across site, Pixel, CAPI and feed, ViewContent and AddToCart events fired via CAPI with content_ids. On Webotic accounts, DPA ROAS lands 2.4 to 3.1x higher than static-creative retargeting on the same audience. The post-ATT fix is to feed the catalog server-side rather than relying on a leaky client Pixel stream on iOS.
How many retargeting audiences should a Moroccan SMB run in parallel?: For 8,000 to 30,000 monthly visitors: 4 to 7 audiences total stacked into two retargeting ad sets at most (one hot 14 d, one warm 30 d), with a hierarchical exclusion stack to prevent overlap. Below 8,000 visitors / month, a single fused ad set, otherwise audiences drop below Meta's learning-phase critical mass. Above 30,000 visitors, add a third ad set "DPA catalog 4-14 d" for e-commerce. Above 80,000 visitors, segment finely by cart value or category.
What's the ideal Meta retargeting frequency cap in Morocco?: 3 to 4 impressions per user over a rolling 7 days, per audience. Past that, CTR drops below 0.8%, CPM creeps up, and negative comments start landing on the creatives. Set the cap via Reach objective when feasible, or via an exclusion rule: drop anyone with 4 impressions in the last 7 days. On CBO or Advantage+ Audience accounts, cap management is exclusion-based only — Meta doesn't expose a native campaign-level cap in those modes.
What's the minimum budget to run retargeting seriously in Morocco?: Retargeting needs a budget proportional to audience size. For a Moroccan SMB with 8,000 to 15,000 monthly visitors, plan 2,500 to 4,500 MAD/month dedicated to retargeting — 25 to 35% of total Meta budget (itself floored at 8,000 MAD to stay in the learning phase). Below 1,500 MAD/month on retargeting, frequency becomes unpilotable — you either over-impression a small audience or under-impression and never exit learning.
How do you cleanly exclude existing customers from retargeting audiences?: Sync your CRM (HubSpot, Pipedrive, Shopify) with Meta via auto-update Custom Audience, hashed on email and +212 phone. Webotic's data.webotic.ma platform runs that sync continuously. Then exclude that CRM audience from every retargeting ad set, on a 60-day minimum window for e-commerce (median Moroccan repurchase cycle > 60 d) and 90 to 180 days for B2B accounts depending on commercial cycle. Commonly forgotten exclusion that removes 12 to 22% of budget waste on scaling accounts.